LRB and 32-bit compressed oops

[Roman Kennke]

>> I don't think that storing fwd ptr compressed is an option. It would
>> have to be decoded and re-encoded in the barrier. It would complicate
>> and slow down everything else in GC. And it would gain us nothing: we
>> still need the 64bits because alignment. (And we already have a
>> prototype to eliminate that word...)
>>
>> The question was really only, can we get rid of the superfluous mov
>> and register usage in the noop-decode case?
> 
> A compressed fwd ptr is a way to get rid of the extra movs. I think it
> would make compiled code better overall, too.


Yeah, but it would disturb a lot of other code. I don't think we want to 
go down that path.

> Without a compressed fwd ptr, messing with the register allocator would
> be required. I have no idea how to do that.

Ok. That might be just fine.

Any idea why it's apparently not a problem with WBs? Shouldn't they 
expand basically the same as LRBs?

Roman